May 20, 2004 at 4:28 pm
Hi everyone,
I have SQL2k on Win2k Advanced Server, all patched up. For past several months, the server would just lock up - the interface - screen, mouse, keyboard would quit responding, the network cards (there are 3 in the system) will drop all connections. Only hard reboot will bring the system back.
I have replaced everything short of mother board including external RAID for Log files and associated SCSI card.
The system (DELL PowerEdge 2400) has 2G of RAM, about 50G available on the RAID where data resides and about 100G on the RAID where Log disks are. The system RAID has about 2G available. These are all separate RAID arrays on separate controllers.
I am running PowerQuest spotlight on it and have enabled performance logging for all vital parameters. Nothing is out of ordinary. The server just locks up. I can't really tie it to any specific stored procedure or transaction. There are about 50 users hitting it through various interfaces - Web, Access, .NET app. There seems to be very little load on the server, but it still crashes at least once daily, at random times. The system has ran in virtually same configuration for past three years.
Has anyone experienced anything like that? Any pointers to where to look or how to figure or what is causing the crashes?
Thanks for your help,
Jakub
May 21, 2004 at 10:29 am
Application Log, System Log, SQL Server Logs.
You can run a SQL Trace, from another machine to see if anything running a horribly written query and killing the machine. Doesn't sound like it if performance counters are stable, but all it takes is on horrible query to bring everything to crawl.
Are we using TCP/IP for all network traffic? Is the server a static IP? Can there be conflict with a work stations ip?
Maybe you are having issues with your router? Do you have any kind of trace software to view the flow of things through your network.
It really sounds like outside influence (IPs or router issues)
If you can't find outside influences from the network or ip managment, burn it down.
This may sound over simplistic but have you run a check disk to see if there is bad sector?
May 21, 2004 at 10:42 am
I just remembered where I saw something like this, I had a client that was assigning ip addresses from the web. Whenever the web went down, everything went down, hard. If your getting this everyday thats not very likely though. I think a bad sector may be somewhere.
Nothing like a server that needs constant attention to spoil vacation plans. I feel your pain.
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply